开云体育

Please note that the images you display are significantly different from the images on the test data. The links I have given have clearly tested the difference between the two（ #1310）. If the accuracy of your test data is the same as your ability to view the image, then these two bad clones will be able to meet your needs.

Hugen

With all due respect, the thread you cite says clearly there is no operational difference among the fully shielded, "better" versions, the "worse" clones, and the worst. No one has presented operational data that distinguishes clone performance at any level. The thread actually says slightly better performance was had from unshielded clones.

The images presented on the data comparison are indeed cut-and-paste images from advertisements but they are identical to the actual VNAs in my possession. I have posted an actual picture of my units as well so you can see the images used for identification are identical.

The data I have published shows no significant difference between the two clones.

So let's review:

1) No one, including you, have published any data or test results demonstrating a significant performance advantage for the "real" unit versus the clones.

2) No one, including you and me, has been able to publish data showing a significant performance difference among the clones.

3) I have stated but not published data (I am happy to do so) where I compared the results of my two clones to a spectrum analyzer/tracking generator/RF bridge combination and found no significant variance in results among all three devices.

The qualities of the original versus the clones...... so far..... have been distinctions without differences. Those differences have been cosmetic only and no one has shown a resulting measurement discrepancy.

I challenge you or anyone on this board to publish actual data showing that a clone performs significantly worse than an "original". I don't think that has been done. The thread you cite certainly does not support measured differences. If you have data please show it. If not then I think we all need to stop confusing the readers here. Some people have actually returned what are likely perfectly good devices because of cosmetics.

WA8TOD

You're testing two bad clones, and the results of the two tests should be consistent, but you're using a good cloned image when presenting the test data, which can be misleading.The group has already discussed the role of shielding./g/nanovna-users/topic/the_benefits_of_screening/33021262?p=,,,20,0,0,0::recentpostdate%2Fsticky,,,20,2,20,33021262
Their difference is not only shielding, but the good isolation of the T/R port can effectively improve the measurement accuracy of the VNA.

It may be worthwhile to attempt soldering the prongs on the other side because it will not only improve mechanical reliability but may also benefit RF performances (however, the RF improvement will only be significant above the nanoVNA operating range).

Whether you decide to solder the bottom prongs or not, connector savers (sacrificial connectors) are still necessary because SMAs have limited mating cycles, especially the cheaper no-brand ones which the nanoVNa presumably come with.

I have been working on what I hope is an easy to follow introduction to the nanoVNA. It's fairly basic and aimed at users who have probably used other test test equipment such as oscilloscope or maybe even a spectrum analyser, but are unsure why they might benefit from a VNA.

This is a work in progress and I will aim to take any feedback given here to improve the content.

Part 2 is in the works and I hope to have it uploaded this time next week.

Has anyone though of upgrading the STM32F072C8T6 to other MCU with larger SRAM and flash memory? The original NanoVNA author, ttrftech, uses STM32F303 for his other project CentSDR and has the chip data defined at



From the datasheet, it seems that STM32F303CCT6 is a good replacement for the original STM32F072C8T6. 256KB flash memory versus 64KB; 40KB SRAM versus 16KB. This enables the possibility of enlarge the sample points from 101 to more points, and other features.

Before I start rework on th LQFP-48 IC for the first time in my life, I would appreciate your feedback on the feasibility and suggestions.

Ken

Jump on in, Reg, the water's fine!

Personally, I would prefer that the TDR function be built into a Windows app (e.g. NanoVNASharp) -- I find using the the NanoVNA's touchscreen more than a little frustrating (e.g. entering frequencies). But that's just me -- each to their own taste.

In lieu of a stand-alone app that incorporates the TDR function, a program could be written to run (probably real-time) under a Python or Matlab environment, for those who have either. (Rune's might be real-time -- I haven't tried any of the Python-based programs).

As for myself -- I'm just writing my Matlab code for the fun of it. (And maybe someone else might find it useful.) And, maybe someday I'll incorporate it into a GUI running under the Matlab environment. But, to be honest, I've been an engineer since '77 and have never had a need for a TDR function. So would I ever use it? Doubtful, but you never know.

- Jeff, k6jca

By the way, if you happen to already know the length of your coax cable but not its velocity factor, with just a few more lines of code you can use the TDR function to calculate the coax cable's actual VF (should one ever need it). Here's what I did, but there is probably a simpler way...

Given (variables in the TDR routine):
Vfactor: Velocity Factor that was input into the TDR routine -- if trying to find actual Vfactor, this can be set to anything (e.g. 0.5).
length_in: TDR calculated length (based on Vfactor), in inches (could be in meters, too)

%
% extra stuff for calculating velocity factor if coax length is known...
% example: my unknown coax of length 164 inches.
%
actual_length_inches = 164; % Yes, I am using inches. That's what my tape-measure measures.
actual_length_meters = actual_length_inches/39.370; % Not used -- just a check.
correction_factor = actual_length_inches/length_in; % divide actual length by TDR calculated length...
vfactor_actual = Vfactor*correction_factor; % and apply as a correction factor

to Frank S,
I am interested in ur "kit" but cannot find ur contact info, sri as I am still noobie lol
pls contact me @ n8aum-at-yahoo-com

I am absolutely blown away by this thread. I was working on TDR via VNA, but had not started to write any code. Just looking at the mathematics very closely. Particularly because of the manner in which the problem is treated in Dunsmore's book, which has a strong "Professionals at work, don't try this yourself." tone.

Clearly the basic work is done, so I am changing direction and will focus on crafting test cases. I spent much of my career in the oil industry finding and fixing bugs in 2.5 million lines of seismic processing codes. The oil industry seismic community was doing DSP almost 20 years before EE started doing it. We could make money with a 125 Hz Nyquist. Digital processing was *much* easier than doing it on analog data. So Geophysical Services Incorporated, the parent company of Texas Instruments, threw huge sums of money at TI to figure out how to make the data acquisition equipment. Eventually TI sold GSI which later shut down for lack of revenue during one of the downturns.

The resolution for TDR work is a function of the maximum frequency and the phase resolution. I don't yet know if we can wedge an FFT into flash or not. I've got a prime factor FFT algorithm which is probably 2 dozen lines of FORTRAN, so it might be possible. But at $35 delivered from China, writing TDR only FW would make it a *very* useful tool worth having for even occasional use.

I've got some chores to do and you're all running *very* fast. I'll try not to fall too far behind, but it's likely to take me a while to catch up. Kirkby talking me into getting one of these when we were chatting on the phone and Dave was right. Wow! It is a tour de force of T&M design.

Have Fun!
Reg

On Sun, 1 Sep 2019 at 15:55, Jeff Anderson <jca1955@...> wrote:

Hi neb (or anyone else who has s1p file for TDR analysis)

Hello

Would it be possible for you to post your s1p file to Github? I've just

written a Matlab script to do the TDR function, and I would like to verify
its math against what others are seeing.

It is obviously good to have code to do the TDR analysis on a .s1p file.
But having the TDR in real time is very useful. To pick the most obvious
example I can think of, if you have *intermittent* problems on a cable, you
can bash the cable around until you see a problem on the time domain. You
can see if there is an open, short, or some other change in impedance that
might indicate that the dielectric is damaged resulting in a change of
impedance.

You can’t do that very easily post processing Touchstone files, bug
obviously it is best to start development of code outside the VNA.

There are several ways I can process .s1p to get time domain data, but
whilst the first two require hardware, so cost money, the others are
available to anyone for *FREE*.

*Expensive as they need hardware *
1) I have access to a Keysight N5247A. I can read a .s1p file and give you
the time domain. That instrument covers 10 MHz to 67 GHz.

2) If you want me to make low frequency measurements down to 300 kHz, I can
do them on an 8753ES VNA. I can save a Touchstone file, but I don’t know if
it even possible to save the time domain data, despite the instrument has
the time domain option.

*FREE for anyone *
3) The software for the relatively low cost 1.3 GHz VNWA can be downloaded
*FREE*. That will work in demo mode without owning a VNWA. It can open a
.s1p file and produce the time domain data.

The VNWA software has more configurable TDR parameters than any other VNA -
more than the PNA-X which I expect cost over a million USD. I am unsure of
the usefulness of all these options on the VNWA though.

4) The software for the Copper Mountain VNAs is also a *FREE* download.
That runs in demo mode without the hardware.

5) The software for the LA Techniques VNAs will run in demo mode. I have
only used that a couple of times, so are unsure of its capabilities.

There are probably other USB controlled VNAs, for which the software can be
downloaded free and run in demo mode. Personally I would only use your time
on the very professional Copper Mountain software or the VNWA. Other
software, (in particular the Picktech VNA) is just a joke.

The good thing about the VNWA software is there’s very good support on the
VNWA forum, including from the developer.

For example, with a length of coax whose S11 was measured from 1 to 60
MHz, I get the attached responses (the time-domain response may or may not
be correct, which is what I would like to verify).

I suggest that you look with the VNWA software. If you get stuck, ask on
the VNWA list.

If anyone would like to do the same thing, my s1p file for this
measurement is here:

Thanks,

- Jeff, k6jca

Dave, G8WRB

--

Dr. David Kirkby,
Kirkby Microwave Ltd,
drkirkby@...

Telephone 01621-680100./ +44 1621 680100

Registered in England & Wales.
Company number 08914892.
Registered office:
Stokes Hall Lodge,
Burnham Rd,
Althorne,
Chelmsford,
Essex,
CM3 6DT,
United Kingdom

Here is the MENU flow chart:

/g/nanovna-users/files/Miscellaneous/nanoVNA%20Menu%20Structure%20v1.1.pdf

This is helpful as it permits you to get an eagle eye view of the instrument soft key flow top view.

Alan

Hi Bruce,

Try this file:

/g/nanovna-users/files/NanoVNA%20Calibration%20Considerations%20and%20Procedure_v1.0.pdf

It is in the FILE DIRECTORY FOLDER.

Also, there is a MENU breakdown flow chart that is very easy to follow and will provide quite a bit of assistance with the manual in hand.

Alan

(forgot to add...)

....And so I'm happy with the results from the Matlab TDR script.

- Jeff, k6jca

Thanks very much, Rune. I was wondering why they weren't exactly the same. And so I compared your code at to my Matlab script (listed, below).

Functionally, your python and my matlab code look identical to me, unless I missed something. (Note: apart from one small difference -- the x-axis of my time-response plot is divided-by-two, so that the x-coordinate of the blip matches the stated length).

Best regards,

- Jeff, k6jca

% ****
% Time-Domain Response from S11 data
% (requires .s1p file) -- K6JCA
% ****

s = 'RG-8_0.66_6.81m_10_282MHz.s1p'; % file containing s11 data
[Gamma_Obj,coax_Notes,coax_State] = spar_read(path,s); % get s-parameter data.

Zo = Gamma_Obj.Z0; % This should be 50 ohms
FreqHz = Gamma_Obj.Freq; % AND assume Freq Vector is the same for all as well.
s11 = squeeze(Gamma_Obj.S_Parameters(1,1,:));

Z_coax = gamma2z(Gamma_Obj.S_Parameters,Zo);
Zin = squeeze(Z_coax);
Rloss = 20.*log10(abs((Zin-Zo)./(Zin+Zo)));

Freq = FreqHz.'; % convert column vector to row vector
FreqMHz = Freq./1e6; % convert Hz to MHz

% Vfactor = 0.69; % velocity factor for RG303 coax
% Vfactor = 0.84; % velocity factor for 9913 coax
Vfactor = 0.66; % velocity factor for RG58A coax

c0 = 299792458; % speed of light in vacuum, m/s

[Row,Npoints]=size(Freq); % Number of points in s11 file (Odd Integer)
start_f = FreqMHz(1);
stop_f = FreqMHz(Npoints);
delta_f = Freq(2)-Freq(1);

window = blackman(Npoints);
windowed_s11 = window.*s11;
NFFT = 16384;
td = abs(ifft(windowed_s11, NFFT));
delta_time = 1 / delta_f;
t_axis = linspace(0, delta_time, NFFT);

d_axis = t_axis * c0 * Vfactor / 2;
[max_td,index] = max(td);
length = d_axis(index);
length_in = length*39.370;
length_feet = length_in/12;
rem_in = mod(length_in,12);

figure(1);
plot(FreqMHz,Rloss);
tlbl1 = sprintf('Return Loss, Frequency Domain Response');
freq_lbl = sprintf('(Start %.3f MHz, Stop %.3f MHz)',start_f,stop_f);
title1 = char(tlbl1,s,freq_lbl);
title(title1,'Interpreter','none');
ylabel('Return Loss (dB)');
xlabel('Frequency (MHz)');
grid on;
set(gcf,'position',[400,500,550,300])


figure(3);
plot(d_axis, td,'linewidth',2);
ylabel("magnitude");
tlbl3 = sprintf('Return Loss, Time Domain');
xlbl = sprintf('length (m)\n\r \n\rLength = %.3f meters,... given Vf = %.3f\n\r(%.2f inches, or %.0f '', %.2f ")\n\r',...
length,Vfactor,length_in,length_feet,rem_in);
title3 = char(tlbl3,s);
title(title3,'Interpreter','none');
xlabel(xlbl);
grid on;
set(gcf,'position',[600,300,550,300])

For those looking for the compatible 3.2 inch display, just do a search on eBay with the following two items in the same line: ili9341 3.2
You'll see individual panels for sale. Check the pinouts in the description and compare to the 2.8" display pinouts (check the nano's schematic for that).

Regards
Larry

It's always bewildering to wander into someplace and find that you're trapped inside
someone's jargon jungle. I've found myself in that shape when I wandered into the
nanoVNA Users group.

I'm a noobie when it comes to the nanoVNA. I've owned a miniVNA but the gui is far different.
The interface presents measurements according to known task names making selections a lot
simpler. The price difference between the two is 10 to 1. When the magic smoke got let out of
miniVNA, I knew I needed a much more affordable instrument. I don't need lab grade for what
I do hence my purchase of the nanoVNA, a good deal.

I came to this group looking for help. I'm sure it's here but filtering through the myriad of messages
make the discovery task a chore. The user manual really isn't too much help. It is obviously written
by someone whose first language is not English but the effort is appreciated. Looking for help in
the various messages means wading through much expert-speak to get to a simple explanation.

What is needed by this oldie-newbie is a manual, or series of tutorials, which give, step by step,
instructions which uses the nanoVNA to perform simple(?) tasks. For instance, proper calibration
is emphasized throughout many messages but nowhere are there instructions which show how to
perform calibration. And if more than one calibration type is required, a tutorial which shows how
to do each.

Something else, there is a document which details the console commands for the nanoVNA. What
application is used to communicate with the device? Surely not the touch screen! And if you
communicate serially through the USB port, what are the parameters of that port?

For other tasks to be performed by the nano, just look at the miniVNA document referenced elsewhere
in this forum. It's a good document but, as I said, the user interface is far different from the nano.

Folks, I'm not complaining (well, maybe I am ... I'm old and in a hurry), but I just want to get along
with my learning so that I, too, can acquire a new expert-speak language, voila'!

Bruce, K4TQL

Wow! Have I got a lot of catching up to do!

I tried, but was unable to find out anything about the xtal which on mine is marked "H26.0 D439" and measures ~3.6 x 2.1 mm.

Between xtal and cable variations on the physical side and sampling on the mathematical side, 0.1% accuracy seems rather good.

I'll create some test cases to run through the TDR codes where we know what the answer should be.

Does anyone know how much of flash is unused on the STM32F072C8T6? It would be *really* useful to have a built in TDR function.

I'm staggered that these are $35 delivered from China. I bought one for $50 because I didn't want to wait.

Have Fun!
Reg

I forgot to screenshot the lengths my software gave, so I'm not sure.
Whoops. But the 8.5m coax is still connected, and currently shows at 8.516
at 0.66 Vf measured to 800MHz. So we're within 8mm of each other.

I doubt we're going to get much better results than this anyway. And the
velocity factor is usually given to an accuracy of "0.66" or rarely "0.659"
- and our difference is equivalent to a difference in Vf of 0.0006.

--
Rune / 5Q5R

Many thanks, neb!

My length calculations are 30.056, 1.24, and 6.814 meters for your three files. Quite close to your length (but close enough?)

Best regards,

- Jeff, k6jca

Many thanks, Rune!

Using your files, my Matlab script gives me the following lengths (assuming Vf of 0.66):

- RG58 measured to 60 MHz: 1.424m
- RG58 measured to 800 MHz: 1.363m
- RG213 measured to 60 MHz: 8.545m
- RG213 measured to 800 MHz: 8.524m

Do you calculate the same?

- Jeff, k6jca